Western blot OD analysis

[Youngmee Sul]

I performed western blot of transfected cos7 cells to determine the levels of our protein of interest. This is our first time using imagej to quantitatively analyze western blot. To test imagej's quantitative capabilities, we loaded 3ul, 6ul, 12ul and 18ul of the same protein in different wells. The gel showed a gradual increase in intensity of the band proportionate to the increase in protein loaded.

I want to determine the densitometry of each of these bands. I used the following procedure:
1. subtract the background with rolling ball radius of 25.
2. outline 1st band (usually start with the brightest band) with box selection tool
3. ctrl + 1
4. outline next band
5. ctrl + 2
6. repeat steps 4 and 5 until all the bands are selected
7. ctrl + 3 (the plots appear)
8. using the straight line tool, I close the bases of all the inverted parabolas.
9. using magic wand, select inside the enclosed parabolas to obtain the areas
10. i used the areas as a measurement of relative intensity.

This method did not yield the appropriate ratios that indicate the correct increase in protein loaded.

What am I doing wrong? Will the areas determined from the gel analysis not reflect the increase in protein loaded? Is there another method to quantify western blots?

[Jos=?ISO-8859-1?Q?=C3=A9?= Gomes]

hello.

i used the image to wb analisys too. however I select all bands in the same
moment in the horizontal line. not each band. because if you select each
band the number genereated is one, if you select all bands in the same time
the number genereated is tottaly different from results obtained from each
band separated. i already observed these differences in my analisys.

i believe that  when you select all bands in the same time the program read
the bands and produced a result based in their interactions and the results
are more coeherent.  try it to you see if your results correspond to that
observed vissually.

good look

Gomes/Brazil

2010/4/16 Youngmee Sul <[hidden email]>

> I performed western blot of transfected cos7 cells to determine the levels
> of our protein of interest. This is our first time using imagej to
> quantitatively analyze western blot. To test imagej's quantitative
> capabilities, we loaded 3ul, 6ul, 12ul and 18ul of the same protein in
> different wells. The gel showed a gradual increase in intensity of the band
> proportionate to the increase in protein loaded.
>
> I want to determine the densitometry of each of these bands. I used the
> following procedure:
> 1. subtract the background with rolling ball radius of 25.
> 2. outline 1st band (usually start with the brightest band) with box
> selection tool
> 3. ctrl + 1
> 4. outline next band
> 5. ctrle  + 2
> 6. repeat steps 4 and 5 until all the bands are selected
> 7. ctrl + 3 (the plots appear)
> 8. using the straight line tool, I close the bases of all the inverted
> parabolas.
> 9. using magic wand, select inside the enclosed parabolas to obtain the
> areas
> 10. i used the areas as a measurement of relative intensity.
>
> This method did not yield the appropriate ratios that indicate the correct
> increase in protein loaded.
>
> What am I doing wrong? Will the areas determined from the gel analysis not
> reflect the increase in protein loaded? Is there another method to quantify
> western blots?
>

--
Jose Rosa Gomes -
Mestre em Ciências pelo ICB-USP-área Histologia e Embriologia
Doutor em Biologia Buco-Dental pela FOP/UNICAMP -área Histologia e
Embriologia
Professor Assistente D do Curso de Odontologia da Universidade Estadual de
Ponta Grossa -www.uepg.br-
Editor Chefe do periódico Animal Biology Journal
https://www.novapublishers.com/catalog/product_info.php?products_id=8943 -
New York - USA

[Wolfgang Schechinger]

Dear Youngmee,

please tell us how you obtained the image you want to evaluate (X-ray film, fluorescence scanner, etc etc)

Wo

-------- Original-Nachricht --------
> Datum: Fri, 16 Apr 2010 19:02:04 -0400
> Von: Youngmee Sul <[hidden email]>
> An: [hidden email]
> Betreff: Western blot OD analysis

> I performed western blot of transfected cos7 cells to determine the levels
> of our protein of interest. This is our first time using imagej to
> quantitatively analyze western blot. To test imagej's quantitative capabilities,
> we loaded 3ul, 6ul, 12ul and 18ul of the same protein in different wells.
> The gel showed a gradual increase in intensity of the band proportionate to
> the increase in protein loaded.
>
> I want to determine the densitometry of each of these bands. I used the
> following procedure:
> 1. subtract the background with rolling ball radius of 25.
> 2. outline 1st band (usually start with the brightest band) with box
> selection tool
> 3. ctrl + 1
> 4. outline next band
> 5. ctrl + 2
> 6. repeat steps 4 and 5 until all the bands are selected
> 7. ctrl + 3 (the plots appear)
> 8. using the straight line tool, I close the bases of all the inverted
> parabolas.
> 9. using magic wand, select inside the enclosed parabolas to obtain the
> areas
> 10. i used the areas as a measurement of relative intensity.
>
> This method did not yield the appropriate ratios that indicate the correct
> increase in protein loaded.
>
> What am I doing wrong? Will the areas determined from the gel analysis not
> reflect the increase in protein loaded? Is there another method to
> quantify western blots?

--
GRATIS für alle GMX-Mitglieder: Die maxdome Movie-FLAT!
Jetzt freischalten unter http://portal.gmx.net/de/go/maxdome01

[Gabriel Landini]

> 1. subtract the background with rolling ball radius of 25.

Applying "a-posteriori" background correction methods on images that will be
subjected to optical density measurements seems to me a very bad idea.

Cheers
G.

[Joel Sheffield]

Gel quantitation is quite complex.  First, you have to be sure that your
image is actually one of transmission.  I have seen students use images
scanned on a flatbed scanner, in which the light passes through the gel and
is then reflected back through the gel again on its path to the detector
--not good!  If you are using a camera to collect the data, you need to know
about the linearity of the camera as well.  You also have to be very careful
about linearity of the stain.  What about saturation of the stain/film?  Is
width/area of the band proportional to the concentration?

Under separate cover, (since this list does not transmit attachments), I am
sending you an exercise that I developed for our cell biology class, using
ImageJ to quantitate a gel.  More than anything else, it indicates the
problems that you might face.  I agree with Gabriel that applying a
background correction is a very bad idea, since you have very little control
of the level.  In this method, rather than isolate specific bands, we scan
horizontally across a concentration series, and then integrate the result,
after adding a "background" line to the graph.

As an aside, one of the interesting approaches to background correction for
an individual band is to calculate the local integrated density under the
edges of the selection box.  This method is used in ImageQuant software.
The problem is deciding on the boundaries of the band.   I can imagine a
macro to do this, but haven't developed one.

Best of luck,

Joel


On Sat, Apr 17, 2010 at 6:15 AM, Gabriel Landini <[hidden email]>wrote:

> > 1. subtract the background with rolling ball radius of 25.
>
> Applying "a-posteriori" background correction methods on images that will
> be
> subjected to optical density measurements seems to me a very bad idea.
>
> Cheers
> G.
>



--


Joel B. Sheffield, Ph.D
Department of Biology
Temple University
Philadelphia, PA 19122
Voice: 215 204 8839
e-mail: [hidden email]
URL:  http://astro.temple.edu/~jbs

[Daniel James White]

Hi Youngmee ,

On Apr 17, 2010, at 6:03 AM, IMAGEJ automatic digest system wrote:

> Date:    Fri, 16 Apr 2010 19:02:04 -0400
> From:    Youngmee Sul <[hidden email]>
> Subject: Western blot OD analysis
>
> I performed western blot of transfected cos7 cells to determine the  
> level=
> s of our protein of interest. This is our first time using imagej to  
> quan=
> titatively analyze western blot. To test imagej's quantitative  
> capabiliti=
> es, we loaded 3ul, 6ul, 12ul and 18ul of the same protein in  
> different we=
> lls. The gel showed a gradual increase in intensity of the band  
> proportio=
> nate to the increase in protein loaded.=20
>
> I want to determine the densitometry of each of these bands. I used  
> the f=
> ollowing procedure:
> 1. subtract the background with rolling ball radius of 25.
> 2. outline 1st band (usually start with the brightest band) with box  
> sele=
> ction tool
> 3. ctrl + 1
> 4. outline next band
> 5. ctrl + 2
> 6. repeat steps 4 and 5 until all the bands are selected
> 7. ctrl + 3 (the plots appear)
> 8. using the straight line tool, I close the bases of all the  
> inverted pa=
> rabolas.
> 9. using magic wand, select inside the enclosed parabolas to obtain  
> the a=
> reas
> 10. i used the areas as a measurement of relative intensity.

all good probably until step 10.

what logic do you follow to assume that the western blot  protein  
detection
should have the ares of the band proportional to the amount of protein.

Surely you should measure the integrated intensity of the signal in  
the area you identify for each band?

how is the signal that is detected generated?
Is it using a kit that makes a dark stain on the white transfer filter  
that you then scan into digital form with a flat bed scanner,
or is it chemo luminescence that is detected with photographic film or  
directly by some imaging device?

you need to measure a signal in the blot that is reasonalby expected  
to be linear with the amount of protein loaded on the gel.

you mention "dosimetry" and "OD"
then say you measure area?

i think you need to think more carefully about where the protein  
amount signal is coming from,
how it is captured digitally, and then how you quantify it.


>
> This method did not yield the appropriate ratios that indicate the  
> correc=
> t increase in protein loaded.=20

i would not expect band area to be linearly correlated with protein  
amount.,
why should it be?

>
> What am I doing wrong? Will the areas determined from the gel  
> analysis no=
> t reflect the increase in protein loaded?

i dont think so.

> Is there another method to quan=
> tify western blots?=20

measure the intensity of the staining or light meaasured added up in  
all the pixels that each band covers.

if you scan in a whit background and want to measure dark patches,  
\you can invert the image to get a positive signal.

good luck

Dr. Daniel James White BSc. (Hons.) PhD
Senior Microscopist / Image Processing and Analysis
Light Microscopy Facility
Max Planck Institute of Molecular Cell Biology and Genetics
Pfotenhauerstrasse 108
01307 DRESDEN
Germany

[phone] (German Mobile)
+49  (0)351 210 2627 (Work phone at MPI-CBG)
+49  (0)351 210 1078 (Fax MPI-CBG LMF)

http://www.bioimagexd.net                BioImageXD
http://pacific.mpi-cbg.de                Fiji (is just ImageJ - batteries included)
http://www.chalkie.org.uk
[hidden email]
( [hidden email] )